Internet of Things requires analysis of big data

To get value from the new industrial revolution that the Internet of Things (IoT) represents, companies must develop tools and skills to turn IoT data into useful insights.

Networks of objects embedded with sensors, electronics and software are here. No longer are sensors limited to capital equipment and factory infrastructure. They are everywhere: from traffic signals that optimize traffic flow to watches that monitor their wearers’ vital signs. The economics of IoT are incredibly promising: Sensors are getting smaller and cheaper, requiring less power and delivering more computing capacity. Meanwhile, the cloud provides ever cheaper storage to hold all the new data. But challenges remain. Companies must identify meaningful, actionable information in the flood of structured and unstructured IoT information. Currently, less than one percent of sensor data (which can be several hundred billion messages daily) is used productively.

Business applications like customer relationship management (CRM) and enterprise resource planning (ERP) have traditionally run on networks that are separate from systems like HVAC and manufacturing equipment. Data interchange protocols are needed to merge disparate data into a common format. Fortunately, industry-specific, cloud-based Software-as-a-Service (SaaS) solutions are available to integrate systems, provide security and perform data analytics. A new generation of data aggregation and analysis tools has emerged. Among them: Apache's open-source Hadoop distributed data processing system and NoSQL databases and business intelligence (BI) platforms like MongoDB. Then there is the organizational challenge. Massive volumes of IoT data will be spread across departments, lines of business teams, and locations in different formats and systems. Time that is spent cleansing and reconciling data reduces time available to analyze, communicate or respond to it.

ICT needs to provide an environment and platform for the full cycle of data analysis and management. This means developing and refining data models, analytics and decision support. Some IoT scenarios require decision support in real-time, while others are “batch” in nature. The technology and business process for these scenarios vary. Most use cases require a hybrid approach that includes both. However, one thing is certain: Big Data analysis techniques must be  applied to IoT data in order to fully realize the impact of the IoT investment. Taking advantage of IoT will require different ways of thinking about how data is used, how much of it you can handle, how fast you can process and analyze it, and where and how decisions based on it are made. Winning companies will overcome these challenges, seizing the step-change in capability that IoT offers to improve performance and connection with customers.